Journal
IATSS RESEARCH
Volume 47, Issue 2, Pages 286-298Publisher
ELSEVIER SCI LTD
DOI: 10.1016/j.iatssr.2023.05.001
Keywords
Centerline rumble strips; Intervention; Forecast modeling technique; Auto-regressive integrated moving average in-tervention model
Categories
Ask authors/readers for more resources
Centerline rumble strips are low-cost effective countermeasures installed on the center of the highway segments to reduce crashes. This study aims to explore Intervention Time Series Analysis approach as an alternative method for the safety evaluation of centerline rumble strips on rural-two-lane undivided highways in Louisiana.
Centerline rumble strips are low-cost effective countermeasures installed on the center of the highway segments to reduce crashes, especially roadway departure crashes. For safety evaluation of centerline rumble strips, methodologies such as naive before-after analysis and cross-sectional study with Empirical Bayes have been widely utilized. The implementation of these methodologies may be limited due to the lack of relevant control groups, and/or other temporal variations in crashes such as seasonality and serial autocorrelation. This study aims to explore Intervention Time Series Analysis approach as an alternative method for the safety evaluation of centerline rumble strips on rural-two-lane undivided highways in Louisiana. Two different methodologies are explored in the intervention time series approach including the Forecast modeling technique and the Auto-regressive Integrated Moving Average intervention model. The forecast models are based on the exponential smoothing technique, state-space framework, and neural network model. The database consists of monthly observations of total and target crashes on 312 highway segments of 1274 miles in length in which centerline rumble strips were installed during the 2010-2012 period. The time frame 2005-2012 is defined as the pre-intervention period whereas the time frame 2013-2017 is defined as the post-intervention period. The analysis revealed that the Auto-regressive Integrated Moving Average intervention model performed better in terms of error estimates including root means square error, mean absolute error, and mean absolute percentage error. The proposed Auto-regressive Integrated Moving Average intervention model reveals a 17.75% total and 40.54% target crash reduction on the selected rural-two-lane undivided highway segments during the post-intervention period. All the findings are found statistically significant at a 95% confidence level.& COPY; 2023 International Association of Traffic and Safety Sciences. Production and hosting by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available